We believe expert generalists are underestimated.
They spend years getting good at something, then find a use for it somewhere unexpected. Their interests can look unrelated until you understand what they are trying to do. By then, the work may already be well under way.
We are building the Expert Generalist Institute to get close to that work early: to know the people behind it, understand where their knowledge gives them an advantage, and help them build.
Our worldview
The problems worth solving rarely fit inside a single discipline. Yet much of education, work and funding still asks people to make sense within one.
We think that mismatch is becoming more consequential. As AI tools become more capable and widely used, we expect more people to be able to attempt work that once required a larger team. A scientist might build software for a problem they know intimately. An engineer might use a method from another field to test an idea that previously stayed on paper.
The person doing the work still has to decide what is worth attempting and whether the result is any good. That takes knowledge of the problem, standards developed through practice, and the willingness to learn what is missing. We expect those qualities to matter more as the tools improve.
We want a world in which people can follow a worthwhile question as far as it takes them. That asks something of institutions too: room to work across fields, people who can understand the attempt, and capital willing to back it before it fits a familiar category.
Knowledge that travels
A field gives you a way to see. Spend enough time in it and you learn which details matter, which shortcuts fail, and what a good answer looks like. That knowledge can become unusually valuable somewhere else.
In 1938, Claude Shannon showed how Boolean algebra could be used to analyse and design electrical switching circuits. Logic gave him a way to reason about machinery. The connection became a method that other people could use.
This is the kind of generalism we care about. Someone learns a subject deeply enough to recognise where it might apply, then does the work of finding out. They have to understand the new setting too. A resemblance is only a beginning.
What recognition misses
A degree, a job title and a familiar career path make a person easier to place. They tell you less about what happens when that person puts several bodies of knowledge together.
Consider someone moving from physics to instrument design to software. Each move might be part of the same investigation. If you judge each in isolation, you can miss the reason for all three.
We think this creates a gap between the ability to do valuable work and the ease with which that ability gets recognised. It is the gap we are interested in. Understanding it takes more than being impressed by an unusual CV. You have to spend time with what the person actually knows.
Our bet is that we can recognise some exceptional people earlier by understanding the connections in their work.
Judgment before consensus
We care about the questions someone asks, the details they notice without prompting, and how their work changes after a failed attempt. Those things tell us where their understanding is deepest.
We also care about taste. Someone with standards of their own can explain why they made a choice, what they rejected, and what still bothers them about the result. That is useful information when the work is too early to judge by its reception.
An instinct about a person is a reason to pay attention. We want to test it against their work and against the judgment of people who know things we do not. Being early means making decisions with incomplete information. It also means being willing to find out that we were wrong.
Why an institute
We are starting with small dinners and conversations in Europe. People bring work they care about and a question they have not resolved. We want them to meet someone who can see what they are attempting and has something useful to contribute.
A conversation might lead to a collaborator, a technical objection, a reference or an experiment. Our job is to stay involved long enough to find out what came of it. A good introduction earns its value in what happens afterwards.
Over time, we want the Institute to accumulate something a single gathering cannot: knowledge of who understands a problem, who has tried an approach, and who might help. Each piece of work should make us better at recognising the next one.
That is why we are building an institute. We want the relationships and the judgment to deepen over years.
When the work becomes a company
Some connections between fields give a founder a way to build something others struggle to reproduce. An instrument makes an expensive measurement routine. A scientific method becomes a tool. Software becomes useful because its maker understands a problem the software industry has barely noticed.
The commercial question is whether that advantage matters to a customer. We are interested in founders who can carry an insight through the difficult work of making a product, earning trust and building a business.
We are developing a fund alongside the Institute to back expert generalists building companies. Our interest starts with the knowledge behind the company: what the founder understands, how they came to understand it, and why it changes what they can build.
We are also developing small grants for experiments and research. Work can deserve support well before there is a business, and some of it will always belong outside one.
We want to be useful early enough that, when there is a company to build, the relationship is already there.